Thursday, April 29, 2010

Imploding Hadrosaurs and Dubious Nomina Dubia

By now, everybody has heard about Jack Horner's attempt to KILL our beloved Torosaurus.

Above: Look out Toro, Horner's coming and he's packing an AK! Photo by Daniel Hendricks, licensed.

Well, I say "our," but I've never really given much of a thought to Torosaurus, beyond my 10-year-old self trying to get the mechanized DinoRiders toy. But some people seem to feel inexplicably attached to this name, often and incorrectly translated as "bull lizard" (it's got big horns, and was ornery like a raging bull!). In reality it means "perforated lizard", in reference to the holes in its frill. And it's not like the animal will cease to exist, it's just the mature, let's say "silver back" stage of Triceratops. Or should we say, Torosaurus is the mature morph, and Triceratops is the immature morph, of Agathaumas. As I mentioned before, Agathaumas specimens lack diagnostic characters in the context of an environment with multiple taxa from which it must be distinguished. But if there's only one ceratopsian in town, the identity of those generic-looking postcrania is narrowed down to one option, and it's no longer a nomen dubium.

Above: Dear E.D., Please stop naming isolated vertebrae and teeth. kthx. Photo by F. Gutekunst pre-1897, public domain.

Ah, the nomen dubium, or "doubtful name." This is an unofficial* designation given by scientists to names whose type specimens can't be classified because they're too generic. Like many dinosaurs discovered by E.D. Cope, Agathaumas was named based on partial postcranial remains. As more fossils turned up, it turned out that its "unique" postcranial features were actually characteristic of a larger clade, and all ceratopsids have nearly identical torsos and limbs. So Agathaumas lost it's standing as a valid name, because (lacking a skull) it couldn't be determined which species of ceratopsid the bones came from. True, the genus name could have simply been used for all ceratopsid species with this type of postcrania, but modern taxonomists don't roll that way, preferring genera to be mostly monophyletic. For years, Agathaumas remained known as a nomen dubium. It probably belonged to either Triceratops or Torosaurus, the only contemporary contenders known from better remains, but since no skull was found with the type specimen, we could never know for sure which one was the junior synonym.

* The ICZN does not officially recognize any such thing as a nomen dubium, and contrary to popular belief, has no rules to the effect that family names can't be based on dubious taxa, etc. Hence Titanosauridae (=Saltasauridae), Hadrosauridae (=Lambeosauridae), Troodontidae (=Stenonychosauridae), Deinodontidae (=Tyrannosauridae), Podokesauridae (=Coelophysidae) Ceratopsidae (=Chasmosauridae) and yes, Ornithodesmidae (=Dromaeosauridae) are all perfectly valid.

That is, of course, unless the number of potential synonyms is reduced to 1. With Triceratops and Torosaurus recognized as one genus, there is no reason to think that the type specimen of Agathaumas and the type specimen of Triceratops don't come from the same species. If and when another genus of ceratopsid is ever discovered from the well-sampled late Maastrichtian beds of North America, this could always be reversed, but for now the default hypothesis must be that only one was present, rather than an extremely common genus with a wide range of individual variation (Triceratops) and one shadowy mystery genus known only from a post-crania identical to Triceratops but hey, maybe it had like 20 horns or something, who knows?

Sarcasm aside, the concept of nomina dubia has become a bit crazy over the years. Jaime Headden has recently been on a mini-crusade to this point on the DML, questioning some pretty well ingrained and yet pretty flimsy concepts of what is and isn't "non-diagnostic". Let's come to our senses and realize that stratigraphic, ecological and temporal considerations also need to be taken into account. Yes, Agathaumas is non-diagnostic relative to, say, Chasmosaurus. But those two were not contemporaries, and barring the use of a time machine, the other ceratopsians that could potentially be synonyms of Agathaumas could not have existed in the same time and place as it did. Except for one, Triceratops, which by rights should go the way or either Brontosaurus (abandonment) or Coelophysis (official conservation).

Above: The dinosaur Rioarribasaurus had its name changed to Coelophysis by mistake. It turns out the original Coelophysis was not only a different species, it wasn't even a dinosaur. The original is now known as Euceolophysis, or "true Coelophysis." Photo by Ballista, licensed.

Which brings me to the promised hadrosaur implosion. Thanks largely to the work of Nicolas Campione, the taxonomy of Late Cretaceous hadrosaurs is finally being untangled after nearly two centuries of confusion. Contrary to conventional wisdom, the short-snouted type species Edmontosaurus regalis appears to be present only in the Campanian Horseshoe Canyon Formation, not the end-Mesozoic Lance, Hell Creek, etc. All the long-snouted "edmontosaurs" in the late Maastrichtian formations can be assigned to either Edmontosaurus annectens (previously Anatosaurus) or the truly duck-billed form Anatotian. Whether or not E. annectens should therefore be kept as Edmontosaurus or re-classified back to Anatosaurus is a matter of personal preference.

Above: The hadrosaur of many names, lately Anatotitan, the once and future Thespesius? Photo by Claire Houck, licensed.

But there's another monkey wrench here. Campione, in his un-published abstracts and talks so far, has been agnostic on the validity of Anatotitan as a separate species. In the past it's been suggested that the "duck-bill" is merely a preservational artifact, a crushed skull. But, there are several specimens that show this feature, and one that appears to be transitional between narrow and duck-snouted forms. On the DML, Greg Paul has asserted that there is only one species of late Maastrictian edmontosaur, and that the flat-billed versions are the mature growth stage. They even fit the pattern of growth seen in the earlier E. regalis populations, reaching the same maximum size, but then also sprouting the duckbill.

So, it looks like Anatotian is about to go the way of Torosaurus. Except, now the newly low-diversity late Maastrictian dinosaur fauna threatens to resurrect another long-dead name. Thespesius is known only from vertebrae. Those vertebrae could theoretically have come from any hadrosaur... if all hadrosaurs were immortal highlanders. In reality, though, there appears to have been only one hadrosaur species in this ecosystem. And in that case, it can have only one name. And that name is Thespesius occidentalis, "western wondrous one." That name is not Trachodon annectens, because I think the type specimen of Trachodon mirabilis is a ceratopsian. But I'm looking into that one.

Above: Some of these teeth may or may not belong to Trachodon mirabilis. From Leidy, 1860, public domain.

To continue crushing the dreams of myself in 1988 and fans of Jurassic Park everywhere, I'll next try to research whether or not Manospondylus is really a nomen oblitum like everybody is assuming (and everyone does simply assume this, because can you imagine if it weren't?). Outlook probably not so good.


  1. Taxonomy drives me nuts. It's like the 33rd level of freemasonry. Or the point when you've paid enough to the Scientologists to learn about Xenu. You learn a bunch about paleontology because dinosaurs are awesome, and it's a real blast, and then you start reading about these kinds of debates and you're like, "Oh. Ugh."

    I'm just scared because "Anatotitan" is my favorite dinosaur name ever, and is my username at about a million websites. That's pretty much it.

  2. "With *Triceratops* and *Torosaurus* recognized as one genus, there is no reason to think that the type specimen of *Agathaumas* and the type specimen of *Triceratops* don't come from the same species."

    I think you're being deceptive here by jumping back and forth between genus and species. If *Triceratops horridus* and *T. prorsus* are both valid species, how can we be sure which one *A. sylvestris* is a senior synonym of? Or does everyone agree now that *T. prorsus* = *T. horridus*? (And what about *Nedoceratops hatcheri*, wasn't it also around at the same time?)

  3. God knows the names Tyrannosaurus and Triceratops aren't going to be abandoned any time soon whatever paleotaxonomists conclude. Too many people are still bitter about losing Brontosaurus. Speaking of Brontosaurus, we need to revive that name somehow. And if someone at the ICZN needs to get into an "accident" for that to happen, then so be it!

  4. I've only been alive almost 16 years, so it won't implode my life if Manospondylus indeed has priority over Tyrannosaurus. Likewise the other examples mentioned in this article.

  5. I don't really mind when it comes to names. If they end up being sunk, so be it. Even so, this is still an interesting (and confusing) topic.

  6. @Brad: I don't think so. As long as there's only one genus present, that genus still needs to be Agathaumas. Unless you split the Triceratops species up into multiple genera, then Ag goes away again. I don't know how all this lumping will effect the various trike/Nedoceratops species, we'll just have to wftp.

  7. @Abyssal: Brontosaurus' type species (B. excelsus) is different from Apatosaurus' type species (A. ajax), so there is nothing in the ICZN to stop you from using it. (Remember, the ICZN only governs what the type is and which name has priority in cases of synonymy -- it gives you complete freedom in determining everything else.)

  8. Right, this is something always overlooked in discussions of taxonomy where Bronto/Apato are used as an example. Everyone seems to agree they're different species, and "genus" is a pretty meaningless concept. Mike Taylor, who resurrected Giraffatitan, did so in part based on the principle that since genera are meaningless, every genus should hold only one species, making the old genus/species binomial into more of a two-part species name. Under this principle, Brontosaurus MUST be brought back unless somebody can demonstrate that it's a growth stage or sexual dimorph of something else. As I mentioned above, this would also save Triceratops (coining a new "genus" name for T. prorosus that is), assuming Horner, Fowler et al. can't demonstrate all Triceratops species are individual variation.

    However, this won't save Anatosaurus/titan/Edmontosaurus from Thespesius, not Tyrannosaurus from Manospondylus (which I can find a few uses of post-1899 so it can't be a nomen oblitum).

  9. @ Matt: You may have a point. I have a hard time thinking of an analogous resolved case at the moment. Is it possible for *A. sylvestris* to be nondiagnostic as a species, but *Agathuamas* to be diagnostic as a genus? Would the animal then be called *Agathuamas horridus*?

  10. @Brad Exactly. That's why we still use (some) family-level names based on taxa not usually considered diagnostic at lower levels, like Ceratopsidae. Though as I mentioned in the post this is very inconsistently applied and seems to change on a whim. I remember when Podokesauridae was in very wide use. Then one study points out the Podokesaurus holotype is lost (not even non-diagnostic, just lost) and everybody switches to Coelophysidae, a name originally based not only on a barely diagnostic type, but one that turned out to be a Silesaurid!

  11. Brad- I agree with you taconomy SHOULD work so that if Agathaumas is determinable at genus level but not species, the taxa are called Agathaumas horridus, A. prorsus, etc.. But there was a real world case of this a couple years back with Mochlodon being the same as Zalmoxes, but indeterminate at the species level. The authors called Mochlodon Zalmoxes sp. in the paper, which I and others think is the wrong way to handle it.

    Matt- In regard to Giraffatitan, the point was that it was not necessarily closer to Brachiosaurus than to other brachiosaurids. But with Brontosaurus, excelsus is closer to ajax than to Diplodocus, Barosaurus, Tornieria, Seismosaurus, etc.. So it's fine for it to also be Apatosaurus, unless you want to resurrect Elosaurus and such.

  12. @Mickey Well, assuming all four(?) currently recognized species of Apatosaurus are actually distinct, then whether or not it's Apatosaurus parvus or Elosaurus parvus, A. excelsus or Brontosaurus excelsus, etc. depends entirely on individual genericometer settings. The only hitch is that breaking up the Apatosaurus band would leave louisae with no published genus name.

  13. On the topic of referral here, I would definately note that if the material is indistinguishable from either, lumping them together is an effective tool for recognizing diversity in an actualistic way. But when it comes to remarking on the actual perspective of diversity, unless the "triceratopsine" species are also lumped, sylvestris as a "species" is just as different as those other species are; one can allow the taxonomy to reflect this, a la Giraffatitan (if it was a different entity from brancai).

    If we argue that a species and a genus are truly unique entities from one another, and that a species is not just a clade of individuals, and a "genus" is a clade containing only species; we can argue then that Agathaumas silvestris (as a single species binomen) is essentially indistinguishable from any other "triceratopsine" so far named. Triceratops would then immediately fall under action to be conserved over Agathaumas under the ICZN, and appeal for action in the PhyloCode's convergent plenary powers would follow. Let's be realistic about people's expectations (Paul and Carpenter's plea to the ICZN to reassign the type of Allosaurus fragilis is only more important here).

    Unfortunately, many old names for fragmentary or incomplete material have fallen under such dissuse that much of it is likely to be referrable to better known, later named taxa, but their disregard and treatment has rendered them unusable.

  14. @qilong: I agree with you except that the names are unusable, which is why I think it's important to draw attention to this subject. Taxonomy is an essentially bureaucratic "science," and is useless without a governing set of rules that everyone implicitly agrees to follow. If people prefer the newer names based on more complete material, they need to be conserved (but hopefully in such a way to avoid Coelophysis or Iguanodon like situations where the name is effectively transferred to a completely different animal).

    Manospondylus, for example, is treated as a nomen oblitum by everyone, except that even if it was not in use as a valid taxon by anyone after 1899, it still can never be a nomen oblitum without an official ICZN ruling. The few people who realize this don't seem to care or be in a position to help. IMO, that's not an ideal state of affairs, and it's one that can be remedied by one short petition.

    Even if Horner et al. can manage to conclusively win everyone over and demonstrate that there's only one genus of "triceratopsine", I have my doubts that they will then follow this with a petition to conserve Triceratops over Agathaumas. We'll end up with a system where nomina oblita don't exist and newer names are simply treated as de facto senior synonym completely arbitrarily.

  15. I just noticed this, as it was in your caption and not the body:

    The original type series for Coelophysis bauri, the neotype for Coelophysis bauri (and original type for Rioarribasaurus colberti) and the type of Eucoelophysis baldwini are all different from one another. While the authors of the latter species did regard a portion of the type series of the first species to potentially belong to it, raising the specter of the purpose of the nomenclature, one can still separate out what is arguably theropod dinopsaur material from the hypodigm and retain it as Theropoda, Coelophysis indet. or something. Ezcurra noted that the specimen referred to Eucoelophysis from Coelophysis, a pubis, does in fact appear to be a coelophysoid element, while the sacrum (the lectotype of Coelophysis bauri) is incomparable to the Eucoelophysis baldwini holotype, as it lacks one, and this makes the two taxa questionably assigned to one another.

  16. Maastrichian End Times phylogenetic discourse is all well & good, but what does it really matter? Genus, schmenus!

    Why not just stipulate that one & only one genus & species of major dinotype did & could exist in the Hell Creek/Lance/Frenchman/Scollard ecosystem & be done with it?

    With the Interior Seaway receding, was there really ecospace for more than one each of tyrannosaurs, ornithomimids, oviraptors, dromaeosaurs, troodontids, big flightless opposite birds or smaller theropod of some dubious kind, ceratopsians, pachycephalosaurs, hadrosaurs or smaller ornithopods of some dubious kind.

    Already it seems that there was no room at all in the North American Latest Cretaceous Inn for any therizinosaur or alvarezsaur, which is not surprising in the former case, as they had grown uncomfortably large in Asia. Nor north of the Kirtland Formation need any sauropod apply, even as distinguished & decorated a veteran as Alamosaurus.

    My apologies to any dinosaur group I might have overlooked & offended. And what was Ricardoestesia, anyway? Everyone says, look, no medium sized theropods in Tyrannosaurdom, but what of all those teeth?

  17. On alvarezsaurids, there was "Ornithomimus" minutus...

  18. Mmm, I still don't understand why there are people who want every genus to be monotypic. I just don't get it, it just makes it meaningless, it is just getting rid of genus and using 2 words for species. Not quite of a good move in my opinion, helps making clades...

    But oh well, taxonomy is just a self-made mess, it just revolves arround itself.

  19. Late replies coming in...

    @John "Why not just stipulate that one & only one genus & species of major dinotype did & could exist in the Hell Creek/Lance/Frenchman/Scollard ecosystem & be done with it?"

    We shouldn't stipulate this of course, but in my opinion low diversity should always be the default assumption. You can't prove a negative, and assuming low diversity puts the onus on those who hypothesize that slight differences are due to taxonomy rather than ontogeny or variation within a population or over time. Assuming high diversity guarantees a messying-up of the nomenclature before any science is actually done to see if it's justified.

    "Already it seems that there was no room at all in the North American Latest Cretaceous Inn for any therizinosaur or alvarezsaur, "

    As Albertonykus already pointed out, alvarezsaurids ("O." minutus) and therizinosaurids (undescribed specimens) are known from Lance-age formations.

    "Mmm, I still don't understand why there are people who want every genus to be monotypic. I just don't get it, it just makes it meaningless, it is just getting rid of genus and using 2 words for species."

    That's exactly why some people want all genera to be monotypic. It's not making genus meaningless because it's already meaningless. But we're stuck with binomials, so we may as well treat them like 2-word species names to avoid all the confusion multitypic genera have caused.

  20. "It turns out the original Coelophysis was not only a different species, it wasn't even a dinosaur."

    Are you sure? And if this is the case, does that mean that the Coelophysis Wikipedia article is wrong in calling it a small Triassic theropod? Please clarify!

  21. @Anonymous: The original material named Coelophysis may indeed not be a dinosaur, though it's too fragmentary to be completely sure. Some of it has been found in analyses to be a silesaurid. The wiki article doesn't need to be changed, however, because the name "Coelophysis" itself was transferred to a different fossil, one that definitely IS a dinosaur. So, coelophysis as cope knew it originally is no longer a dinosaur, but the Coelophysis specimens known during the 20th century are. It's a bit confusing, I know!

    (An alternate solution to this problem, of keeping the dubious, possibly non-dinosaur as Coelophysis and re-naming the definite dinosaur we all know to "Rioarribasaurus", was rejected by the official organization that controls scientific names).